Machine-learning multi-scale mapping of outdoor thermal comfort in Freiburg (HTC-NN)
Simulation4.2 GB
Data, submodels and results of the HTC-NN neural network, which maps outdoor human thermal comfort and its intra-urban variability at high resolution. The raw sensor network data must be requested from the University of Freiburg.
Code for HTC-NN and its submodels, which downscale and predict intra-urban microclimate variables and outdoor human thermal comfort at high resolution. Useful as an ML surrogate for urban microclimate.
Harmonised data from a street-level network of 42 automatic weather stations in Freiburg for 2022. Useful for intra-urban temperature and wind variability.
A Python package that combines crowdsourced Netatmo weather-station data with gridded urban morphology, using multi-group PCA and machine learning to quantify urban heat island intensity in Lyon, Toulouse, Rennes and Dijon.
1 m land cover rasters for Freiburg covering buildings, paved surfaces, evergreen and deciduous trees, low vegetation, soil and water, with and without tree canopy. Built from OSM, Urban Atlas, building outlines and lidar as input for urban climate models.
From Zenodo, added 2026-11-15.
Listed there as “Data HTC-NN”. Tags proposed by AI and not yet checked by hand. Spotted a mistake? Tell the portal contacts.